Evaluation of end-of-season drought stress tolerance in some rice genotypes, north of Iran

Drought stress has been well documented as an effective parameter in reducing crop production. Hence, developing and releasing new genotypes adapted to water deficit conditions can be a constructive way to overcome to unsuitable environmental conditions. Canonical discriminant analysis (CDA) was used in combination with cluster analysis to evaluate the response of 14 rice genotypes using the drought stress tolerance indices under two conditions of irrigation and discontinued irrigation at a specific time. The CDA results revealed that the first canonical variable was under the influence of high coefficients for the stress tolerance index (STI), harmonic mean (HM), mean production (MP), geometric mean production (GMP), yield in normal conditions (Yp), and yield in stress condition (Ys).The canonical coefficients of the stress sensitivity index (SSI) and tolerance index (TOL) were significant in the second canonical variable. Accordingly, the first canonical variable distinguished the genotypes based on yield potential and stress tolerance, while the second one segregated the susceptible and tolerant genotypes. The scatter plot chart of two significant canonical variables distinguished three groups, and all pairs of Mahalanobis distances between groups were significant. The second group was recognized as the best group, since its genotypes had the maximum value for the first canonical variable, while most of these genotypes had low values for the second one. Therefore, the genotypes of this group (831, Fajr, Sepidrood and Sahel) can be selected as suitable genotypes for stress and non-stress conditions.